Substantial variation in surgical mortality rates and costs per surgical episode exists across providers, suggesting opportunities for improvement. While a variety of initiatives have been launched to enhance the safety, quality, and cost-efficiency of surgical care, these efforts focus almost exclusively on the surgeon's activities in the operating room and immediately afterward. However, there is growing recognition that coordination of care and communication among the surgeon, the referring primary care physician (PCP), and consultants are crucial. As such, optimizing surgical care will require the ability to measure and understand the care delivered by all physicians involved in the surgical episode and their interactions. In this context, the candidate (Dr. John M. Hollingsworth) seeks to characterize and evaluate the effectiveness of physician social networks that deliver surgical care. The candidate, a urologist and health services researcher at the University of Michigan, will leverage this proposal to broaden the scientific and clinical scope of his research agenda, and to facilitate his development into an independent investigator. During the period of support, he will pursue additional didactic instruction in several disciplines, including graduate- and doctoral-level courses in network theory, health economics, and healthcare organization and policy. He will also have ample opportunities for mentored, project-based learning, including the hands-on application of advanced statistical modeling and the techniques of social network analysis. The research plan has 3 Specific Aims. Specific Aim 1: To map physician social networks that deliver surgical care. With files from Blue Cross Blue Shield of Michigan and Michigan Medicare plans (2008-2010), the candidate will identify beneficiaries undergoing 1 of 4 common inpatient surgical procedures. By surgery type, he will use a validated method to graph physician social networks at hospitals where these procedures are performed, characterizing them in a number of directions across a range of local and global network properties. He hypothesizes that, compared with physician social networks serving rural areas, those at urban hospitals have a larger number of physicians, but their average distances are similar. He posits that physician social networks that care for female and black patients will be sparser than those for male and white patients, respectively. Specific Aim 2: To evaluate the effect of physician social networks on surgical care safety and quality. The candidate will then examine the association between physician connectedness and a hospital's safety culture/teamwork climate. He will also assess for a relationship between measures of surgical quality (i.e., operative mortality, readmissions, and postoperative complications) and local network properties across 3 phases of the surgical episode. He hypothesizes that networks in which the surgeon, referring PCP, and consultant tend to cluster together foster environments of safety. He posits that while surgeon centrality confers lower morbidity and mortality during the operative phase, greater surgeon-PCP tie strength leads to fewer complications and readmissions in the transition phase. Specific Aim 3: To examine the impact of physician social networks on the costs of surgical care. Finally, the candidate will quantify total expenditures for patients cared for by networks with varying levels of physician connectedness. He will explore specific sources of savings by comparing component payments over global network properties. He hypothesizes that significant variation in total episode expenditures exists between networks. He posits that cohesive and dense networks accrue savings from a variety of sources, including lower payments for post-acute care. Completion of the proposed research will inform the Centers for Medicare and Medicaid Services and other large payers on the extent to which strengthening interactions between physicians, in general, might impact on the safety, quality, and costs of surgical care. More broadly, findings from this proposal will provide researchers, clinician leaders, and policymakers with additional insight into the nature of relationships among different types of physicians and their importance for healthcare delivery.